Create Videos from Text, Frames, and References

MiniMax H3 AI Video Generator

Create 4–15 second videos from text, frames, and multimodal references. Guide motion, camera, and visual style with output up to 2K.

Generate a Video View Examples

50 Starter Credits

Text-to-Video

First & Last Frame Control

Image, Video & Audio References

4–15 Seconds · Up to 2K

Create with MiniMax H3

Create Videos with MiniMax H3

Generate 4–15 second videos from text, a required first frame with an optional last frame, or multimodal references. Choose 768P or 2K output and customize your generation settings.

AI Model
Upload an image, video, or audio reference.
Reference materialsImages, videos, and audio combined0/12
Reference Images(max 9)0/9
Reference Videos(max 3, total 15s)0/30.0s/15s
Reference Audios(max 3, total 15s)0/30.0s/15s
Resolution
Aspect Ratio
Reference workflowMulti-Reference GuideView guide

MiniMax H3 accepts text, image, video, and audio references. Combine supported materials to give the model more precise creative direction.

Prompt Guide

Type @ to quickly insert uploaded files into your prompt.

Select a file reference such as @image1, @video1, or @audio1 so MiniMax H3 knows which material to use.

Material Limits

  • Up to 9 reference images
  • Up to 3 reference videos, 15 seconds total
  • Up to 3 reference audio files, 15 seconds total
  • Up to 12 materials across all reference types

Supported Input Combinations

TextText + ImageText + VideoText + AudioText + Image + VideoText + Image + AudioText + Video + AudioText + Image + Video + Audio

Note: Reference images, videos, and audio are optional. Add only the materials needed for your prompt.

Supported file requirements
  • Images: JPG, JPEG, PNG, WEBP, HEIC, or HEIF; up to 30 MB each
  • Videos: MP4 or MOV; 2–15 seconds each and up to 50 MB
  • Audio: WAV or MP3; 2–15 seconds each and up to 15 MB
  • Image and video dimensions: 256–5760 px; aspect ratio 0.4–2.5
  • Total request size: up to 64 MB

Creative starting points

MiniMax H3 AI Video Generator Examples

Watch MiniMax H3 AI video examples built from text prompts, frame controls, and multimodal references across ads, games, and creative work.

Flexible creative controls

Key Features of MiniMax H3

Explore MiniMax H3 features for text-to-video, first and last frame control, 2K output, reference media, motion, camera, and visual style.

Generate from Text

Describe the subject, action, setting, lighting, and camera movement. MiniMax H3 uses your prompt to generate a short video.

Set the Starting or Ending Frame

Upload a starting frame, an ending frame, or both. These images help guide how the video begins or ends.

Add Reference Media

Add supported images, video, or audio to provide more context for the subject, motion, camera movement, and visual style.

Guide Motion and Camera

Describe pans, zooms, tracking shots, static shots, and other camera directions in your prompt.

Create 4–15 Second Videos in Up to 2K

Choose a duration from 4 to 15 seconds and generate the final video in 768P or 2K.

Guide Subject and Style Details

References provide more context for important visual details. Results vary by source quality and scene complexity.

Four-step workflow

How to Use MiniMax H3 Online

Choose a generation workflow, add your prompt and reference media, then customize the settings and generate your video in up to 2K.

01

Create an Account

Create your account and use your starter credits to begin generating videos.

02

Choose a Workflow

Select text-to-video, image-to-video, frame-controlled, or reference-based generation.

03

Write a Prompt and Add References

Describe the video you want to create, then add frames or reference media when you need more control.

04

Generate and Download

Choose a 4–15 second duration and your preferred resolution. Review the credit estimate, generate your video, and download the result.

Example prompt: A silver watch rests on dark stone. Soft light moves across the surface as the camera slowly pushes forward. Premium product commercial style.

Choose the right input

Which Video Workflow Should You Use?

Compare MiniMax H3 workflows by the input you have and the level of control you need.

Workflow

Text-to-Video

A written prompt

Exploring a new scene from scratch

First and Last Frame Video

A prompt and one or two frame images

Animating a still image or guiding a transition

Multimodal Reference Video

A prompt with supported images, video, or audio

Adding context for the subject, motion, camera, and visual style

Practical workflow guide

Best MiniMax H3 Workflows

After you choose an input mode, use these workflows to plan a cleaner video. Pick the job first. Then give MiniMax H3 one clear motion target.

Product Ad Workflow

Best for ecommerce and SaaS launches

  • Use Image to Video when the product shape, color, or logo area must stay stable.
  • Start with a clean product image, simple background, and one camera move.
  • Ask for slow rotation, push-in, or light background motion. Avoid stacking actions.
  • Review logos, labels, edges, and small text before publishing.

Character Motion Workflow

Best for game, anime, and creator concepts

  • Use Image to Video when the face, outfit, or hairstyle must stay consistent.
  • Describe one primary action such as a head turn, hair movement, walk cycle, or pose shift.
  • Keep camera language simple: locked camera, gentle handheld, or slow push-in.
  • Do not combine full-body motion, expression changes, and scene transitions in one short clip.

First and Last Frame Workflow

Best for planned transitions

  • Use this when the opening and ending frame are both important.
  • Write the prompt around the transition between the two frames.
  • Keep the motion path clear: reveal, approach, open, transform, or settle.
  • Use fewer scene details so MiniMax H3 can connect the frames cleanly.

Reference Media Workflow

Best for style, pacing, and camera direction

  • Use references when text alone cannot explain the movement or look.
  • Tell MiniMax H3 what to follow: style, camera, motion rhythm, lighting, or pacing.
  • Do not ask one reference to control everything at once.
  • If the result drifts, keep the strongest reference signal and regenerate.

Simple rule: start from an image when consistency matters. Start from text when the idea matters more than exact identity. Use first and last frames when the ending matters.

Real project workflows

MiniMax H3 Video Generator Use Cases

Use MiniMax H3 to explore product ads, social videos, film shots, game concepts, and motion design before committing to full production.

Product Marketing

Marketing & Product Teams

Turn a product image into a short ad concept before planning a full product shoot.

Ecommerce Content

Ecommerce Teams

Turn existing product photos into short lifestyle video concepts for storefronts and campaigns.

Social Media Videos

Creators & Social Teams

Explore short-form video concepts for Reels, Shorts, TikTok, and other social channels.

Film Previsualization

Filmmakers & Creative Teams

Test shot direction, framing, movement, and scene mood before filming.

Game Concept Development

Game & Concept Teams

Turn character art into short motion and environment studies for early game concepts.

Motion Design

Motion & Brand Designers

Explore animated poster, campaign, and visual identity concepts with controlled motion and style.

Review generated text, logos, packaging, faces, and other brand details before publishing or using the final video.

Plans and credits

Choose a MiniMax H3 Credit Pack

Create 4–15 second MiniMax H3 videos for 9 credits per second at 768P or 14 credits per second at 2K. Buy credits once—purchased credits never expire.

Starter

$9.90one time

370

Credits

Up to 5 five-second 2K videos

Creator

$29.90one time

1,300

Credits

Up to 18 five-second 2K videos

Best value

Studio

$49.90one time

2,500

Credits

Up to 35 five-second 2K videos

Business

$99.90one time

5,550

Credits

Up to 79 five-second 2K videos

Included with every credit pack

  • One-time purchase, no subscription
  • Purchased credits never expire
  • Up to 2K video output
  • Watermark-free downloads

Compare credit usage, billing details, and all available plans on the MiniMax H3 Pricing page.

Frequently asked questions

MiniMax H3 FAQ

Get answers about generating MiniMax H3 videos, choosing inputs and resolution, using reference media, and understanding credit usage.

What is MiniMax H3?

MiniMax H3 is a multimodal AI video generation model that supports text, image, first/last-frame, and reference-based workflows, with 768P or 2K output.

How do I use MiniMax H3 online?

Create an account and choose a workflow. Write a clear video prompt. Add supported reference media if needed, review the credit estimate, and generate your video.

Can MiniMax H3 create videos from text?

Yes. Enter a prompt that describes the subject, action, setting, lighting, and camera movement. MiniMax H3 uses it to generate a 2K video.

Can MiniMax H3 turn an image into a video?

Yes. Use an image as the first or last frame to guide the video, or use images as reference media in the Reference-to-Video workflow.

Can I use both a first frame and a last frame?

Yes. You can upload both frames. They help guide how the video begins and ends.

How long can a MiniMax H3 video be?

Choose a video length from 4 to 15 seconds.

Can MiniMax H3 generate 2K videos?

Yes. Choose 2K for higher-resolution output or 768P for a smaller video output in the current generator.

What can I use as reference media?

You can add supported images, videos, and audio files. Reference media provides more context for the subject, motion, camera movement, and visual style.

What image, video, and audio formats are supported?

Images: JPG, JPEG, PNG, WEBP, HEIC, and HEIF. Videos: MP4 and MOV. Audio: WAV and MP3.

Can I use audio as a reference?

Yes. Use reference audio with a text prompt, either on its own or alongside supported reference images or videos.

How many credits does MiniMax H3 use on this site?

On our platform, MiniMax H3 generation uses 9 credits per second at 768P or 14 credits per second at 2K. A five-second output uses 45 credits at 768P or 70 credits at 2K.

How are credits calculated when I upload a reference video?

Add the total reference video duration to the output duration, then multiply by the selected resolution rate. Three reference seconds plus a five-second output uses 72 credits at 768P or 112 credits at 2K.

How many free credits do I get when I sign up?

You get 50 credits when you create an account.

Is this the official MiniMax website?

No. This is an independent third-party video generator. It is not affiliated with, endorsed by, or operated by MiniMax.

Turn Your Idea Into a Video with MiniMax H3

Start with text, frames, or multimodal references and create a 4–15 second video in up to 2K.

Generate a Video View Examples